Heart and Soul nebula - Moonlight Gradient on Mosaic - Forums

Topic starter

Sep 19, 2022

Hello there,

I have just completed my first Mosaic panel. I chose the Heart and Soul Nebula.
I am looking for any recommandation to improve this first processed try.

I am struggling with the moonlight gradient of the first panel (left one on the Heart nebula). It was shot during full moon (yeah, I like to live dangerously).
I though I would be able to correct this on post-processing (with DBE), but I was not. If you have any idea, I'm listening

If there is no post-processing solution, I guess I will re-shoot this panel.

This might be also oversaturated.

Using Pixinsight for post-processing.

focal 342mm / 3 panels / f.4,5 / 3-4h per panel
Color Camera with Quadnarrowband filter

Thank you for your help !

andrea tasselli

Frequent contributor

Sep 19, 2022

What process are you using for masaic composition?

Alban Lechatellier

Topic starter

Sep 19, 2022

@andrea tasselli
1- Stacking and DBE for each panel
2- ImageSolver
3- Mosaicbycoordinates
4- GradientMergeMosaic (Overlay+6 shrink radius+50feather)

Rob Foster

Sep 19, 2022 · Edited: Sep 19, 2022

Alban, I am relatively new to constructing Mosaics as well. I follow your same general work flow. Check out David Ault's (Trapped Photons) useful link http://trappedphotons.com/blog/?p=994. I have found his DNA Linear Fit script very helpful at helping with these types of gradients. The link provides the ideal work flow and it really worked a lot better than serial before and after DBEs and certainly better than Linear Fit - Let us know if it helped or not!

Helpful Engaging Supportive

andrea tasselli

Frequent contributor

Sep 19, 2022

Use the script PhotometricMosaic.

Tom Boyd

Sep 19, 2022

andrea tasselli:
Use the script PhotometricMosaic.

I second this suggestion. PhotometricMosaic is amazing. I've use it BEFORE DBEing each of the subframes and it handled the gradients extremely well...

Steve Spring

Sep 19, 2022

I'm by no means we'll practiced at mosaics but am loving the resolution gained by it and as usual, finding the scripts others have shared incredible.
My work flow for Mosaics is as follows and has so far provided pleasing results.

1- Stacking each panel
2- ImageSolver
3- Mosaicbycoordinates
4- MosaicTrim
5- PhotometricMosaic to combine
6- ABE
7- Autocolor
8- Denoise
9- Decon if it suits the target
10- Stretch

There are three different methods to joining the panels, Average, Blend and Overlay. One of these may produce better options to mitigate the gradients in your first panel?

Helpful

Alban Lechatellier

Topic starter

Sep 20, 2022

@Steve @Tom Boyd @Rob Foster @andrea tasselli

Thank you for your help !

The PhotometricMosaic process was definitly a game changer. The moonlight quite disappeared, thats just magic.
I do not feel any loss in the signal of the nebula.

See the result below, I have just applied a soft stretch. This is fantastic. Again, thank you so much.

Clear skies !

Respectful Supportive

Oct 29, 2022

Have you attempted to use drizzled data with MosaicByCoordinates?

I have two drizzled (2x) panel mosaic panels of the Veil Nebula. I've successfully ran ImageSolver on both unstretched/linear panels. When executing MosaicByCoordinates the process gets to "High Quality Image Warping" and then simply 'stops' without creating the two registered image files. My output directory is set and accurate.

Is MosaicByCoordinates incompatible with drizzled data? I thought it could be specific to my target and processing, so I tried it on another drizzled target’s panel and the exact same result occurred.

Any ideas?

TIA,
Jason

Well Written Engaging

Alban Lechatellier

Topic starter

Oct 29, 2022

Hello Jason,

I have never tried yet with drizzled data.
I see no reason why this process won't work for you. My guess is that your computer might be struggling with such big data.

Do you have sufficient memory (RAM) ?

Have you checked the pixinsight console for any error ?

Oct 29, 2022

Alban Lechatellier:
Hello Jason,

I have never tried yet with drizzled data.
I see no reason why this process won't work for you. My guess is that your computer might be struggling with such big data.

Do you have sufficient memory (RAM) ?

Have you checked the pixinsight console for any error ?

I have an M1 Macbook Pro with 32GB of RAM. The output on the console simply just “ends” without an error code.

Here is my post on the Pixinsight forum with the contents of the console output:

https://pixinsight.com/forum/index.php?threads/drizzled-panels-and-mosaicbycoordinates-stop-or-crash.19479/

Steve Spring

Oct 29, 2022

Hi Jason,

Could it be the file path lenght. I have issues before, may be worth taking copies and moving them some where closer to your root folder.
Otherwise, check you pixel size settings in MosaicByCoordinates against your fits headers in case theres somehow a mismatch.

Just thoughts - not saying that's it?

Oct 29, 2022

Steve:
Hi Jason,

Could it be the file path lenght. I have issues before, may be worth taking copies and moving them some where closer to your root folder.
Otherwise, check you pixel size settings in MosaicByCoordinates against your fits headers in case theres somehow a mismatch.

Just thoughts - not saying that's it?

I tried moving the two panels to my Users directory (less than 10 characters in the file path) and it resulted in the exact same way.. the script simply stops. I've not been able to confirm MosaicByCoordinates works with drizzled data. I tried using the master light panels (non-drizzled) in MBC and it worked great. I tried using MBC with other drizzled data and it 'stops' in the same place.

Thanks for the suggestion.

Jason

Helpful Concise